The present study investigates stability and variability as distinct features of state self-esteem change. The amount of variance attributable to stable individual differences, the association between successive self-esteem experiences, and situational and/or person × situation interaction effects were estimated in repeated measures of self-esteem whilst considering measurement invariance. N = 160 adolescents reported momentary self-esteem in an experience sampling study, (i.e., three assessments a day, over a 2-day period). Latent state-trait analysis indicate that change in state self-esteem is a stable process largely determined by interindividual differences in trait self-esteem, while the impact of occasion-specific influences is small. These intrinsic dynamics ensure consistent self-referential experiences. Results reveal experience sampling as a suitable method to track short-term changes of self-esteem.